There are two major technical fields that have shown explosive growth over the past few years: the first is wireless communications and the second is use of data services, particularly the Internet. These two technical fields both require a set of specialized tools in order to measure their quality of service. Interestingly, wireless communications and data services are beginning to converge. Unfortunately, this convergence has not been accompanied by the development of appropriate specialized tools to measure data quality of service in the wireless network.
The growth of wireless communications has been astounding. Twenty years ago, there was virtually no use of wireless communications devices such as cellular phones. In contrast, the market penetration for wireless devices in the U.S. in 1999 was 32 percent. The current forecast is that 80 percent of the U.S. population will be wireless subscribers by 2008.
There are a variety of specialized tools that are used to measure quality of service over wireless networks. These include the following (just to name a few examples):                Ascom QVoice (including QVoice unattended);        Ericsson TEMS, RSAT-2000, Benchmarker, CellAD, and CeNA;        Nokia TOM;        Safco VoicePrint, DataPrint, and WalkAbout;        Comarco BaseLINE and Gen II;        Grayson Surveyor;        ZK CellTest DX136 and DXC;        Ameritec Swarm;        Neopoint Datalogger; and        Qualcomm QCTest Retriever and QCTest CAIT.        
The general deficiency with these tools is that they were primarily developed to measure voice quality and/or RF parameters over the wireless system and not to measure data quality. Some of them have been modified to include some rudimentary data measurements; however, they are not optimized for performing wireless data measurements. In particular, they do not allow unattended measurement of wireless data from multiple remote units in a statistically significant manner with remote control from a back end processor.
The classical way of measuring voice quality of service and/or RF parameters in a wireless network involves sending out technicians to drive test the network. The drive test includes placing the test instrument in a vehicle and running a test script that either generates or receives a voice test signal. The receiving end of the communication link uses a DSP containing a model of human hearing to analyze the received voice sample and produce an associated quality score. In addition, some of the systems measure other system parameters such as SINAD, noise, distortion, received signal level, and call progress statistics.
Unfortunately, the classical method of measuring voice quality of service and/or RF parameters does not function very well for measuring data quality of service. In order to make statistically significant measurements of data quality of service over a wireless network, it is necessary to make multiple measurements from multiple remote devices. Furthermore, a measurement of data quality is inherently different from the other types of measurements due to the effects of latency and other effects that are specific to data.
Most of the existing measurement devices do not have this capability for a variety of reasons. The price of the test instruments range anywhere from $5K to $100K. This makes it price prohibitive to field a statistically significant fleet of remote devices. Thus, what is needed are remote devices designed for unattended operation that is remotely controlled by a back end processor in order to reduce manpower costs. Additionally, what is needed are remote devices that are optimized for performing measurements that are useful over wireless data networks, such as latency for Web page access or delay in SMS message delivery.
The growth of data services has been just as astounding as the growth rate for the wireless industry. The largest driving force behind the growth of data services has been the enormous growth of the Internet. For example, there were 130 Web sites in June 1993, 230,000 Web sites in June of 1996, and 10 million Web sites at the end of 1999.
There have been a variety of specialized tools developed to measure the data quality of service over the Internet.
U.S. Pat. No. 6,006,260 to Barrick, Jr. et al. (assigned to Keynote Systems, Inc) discloses a method for gathering latency experienced by a user over a network. The steps of the method include a user browser sending a GET command to retrieve an HTML page with an embedded Java script. The Java script starts a timer and generates a GET command to retrieve an HTML page. When the page is received, the timer is stopped and the timer information along with cookie data stored on the browser machine is sent to a relay server that logs the information.
U.S. Pat. No. 5,657,450 to Rao et al. teaches the provision of time estimates for long-running distal source access operations using an intermediate server close to the client workspace.
U.S. Pat. No. 5,796,952 to Owen et al. discloses a method for monitoring a user's time of page browsing.
U.S. Pat. No. 6,012,096 to Link et al. teaches a method for monitoring client-to-client network latency for gaming applications. The method involves a ping, response, and response-response protocol.
Unfortunately, none of these patents teach a method which is appropriate for performing data quality of service measurements over a wireless network.
As previously mentioned, there is a tremendous convergence taking place that combines the wireless network with data services. Dataquest estimates that the U.S. wireless data market (including phones, PDAs, laptops, and the like.) will grow from 3 million subscribers in 1999 to 36 million subscribers in 2003. Ericsson is estimating that 1 billion wireless units will be in use worldwide by 2003 and that 40 percent (400 million) of these units will be employed by data users. Furthermore, Ericsson is predicting that 2003 will be the crossover year in which wireless Web access will exceed wired Web access.
As a further measure of the explosive growth of the convergence of the wireless systems and the Internet, one can look at projections for the number of wireless portal subscribers. According to the Strategis Group, the number of wireless portals will increase from 300,000 in 2000, to 9.8 million in 2003, and finally to 24.8 million in 2006.
A variety of technical advancements have accelerated the convergence of Internet access over wireless devices. In 1997, three competing handset vendors (Nokia, Ericsson, and Motorola) and a small software company (Phone.com, formerly Unwired Planet) joined forces to create a standard way to transmit Internet data to wireless phones without occupying too much bandwidth. The result of this collaboration was development of the wireless application protocol (WAP). One basic component of WAP was development of the WML (Wireless Markup Language, replacing the previous Phone.com Handheld Device Markup Language, HDML) that compresses Web content in comparison to HTML. Additionally, the WAP forum developed standards for the use of microbrowsers in mobile devices.
Unfortunately, the development of wireless Web access technology has significantly outpaced the development of wireless data measurements tools. Accordingly, there is a tremendous need for specific test tools to address the converging technologies of wireless systems and data communications.